[BERT] Moving Cities and Improvements

[vorlon_mi]

I've been working on a game as NSA to get the "Terror from the Deep" achievement -- founding all aquatic cities. Thus, I've been focusing on moving cities, including how often, and having workers around to improve the hexes after the move.

Insight 1 -- many of you may have figured this out already. When a city moves onto a space, it wipes out the improvement that is already there. In my first several games, I had assumed the opposite; I worked to be sure to improve the space before the move, assuming that the city would "hover" over it. Wrong. Appears that I have been wasting worker turns making improvements that will be squished.
That leads me to do a bit more planning. If I plan to move a city onto a strategic resource, I need to plan to move it again, so that I can get the resource back by improving the tile after the second move. I'm really enjoying the NSA ability to shorten the turns required for moving a city.

Insight 2 -- Buying vs. Moving for acquiring nearby tiles. If I've done a good job in picking the initial city location, I may only need to move twice to get access to all the nearby resources. I then balance the cost of buying a second ring tile (to get floatstone or a food bonus) with energy against the turns/production to move the city again. If I move too often, especially in a linear fashion, I have trouble working the tiles in the city's original footprint. The costs of buying tiles goes up the more often I do it, so some thought is required.

Insight 3 -- Production for Aquatic Cities. Manufactories must be build on land, not water. Thus, I have been building a lot of production-focused buildings and chose the personality trait where city production yields are buffed. What has been working well in this game is to found the city right on the coast of an island, so 2 or 3 tiles in the initial footprint are land. Subsequent moves take the city further into the water, towards resources, but I still have access to any land resource from the initial position.
As I neared my desired victory condition, I moved my most productive city back towards the coast, so that I could have a coastal land hex to build the victory wonder upon.

 

[legalizefreedom]

I will share my strategy. It's a bit different with your starting in water and self-imposed restriction of having only water cities since most sponsors start on land. This gives you a traditional capital to support your water cities when you are ready to settle them. So I won't profess to have a strategy for that, but I'll share my general strategy.

Once I settle a water city, first on the list is a Trade Depot. If I can, I'll have a trader ready for it by the time the building is done. I also prioritize sending trade routes to the new cities so they can get the production kick-started.
Then I build some production buildings before moving it. If there is a more immediate need to get more tiles to improve I may make an exception, but not often. Presuming you have the tech for them, I build:
Trade Depot
Recycler
Drydock
Water Refinery
Drone Command (if available)
Thermohaline Rudder

A bit later in the game, by the time these are all in play, you can usually buy 2-4 of them. This will likely make your Trade Depot take about 2-3 turns. Also if you are the NSA, you can probably leave off the rudder.
Once these are built or I can otherwise move the city in 1 or 2 turns (thanks also to trade routes) I start moving the city and usually don't stop until the entire workable area is done.
Pop out to the second ring and make your way around the second ring until your workable area is done and then slide back to the center and start building anything you're heart desires. Top of the list being defense buildings and ships. You'll also want to strategically pick the right cities for the wonders that give bonus "per coast tile", etc. Those can be really powerful.

There are exceptions, but in general the only time I buy tiles in a water city is if I need a couple inland or if it would take too many moves to get to 1 or 2 hexes that are somehow obstructed. It's also another solution for needing a couple of hexes for my worker to improve instead of interrupting the build queue.

 

[vorlon_mi]

Thanks, @legalizefreedom . This game is unusual, intended to get the Steam achievement. Normally, I do settle a mixture of land and water cities, to get resources and grow my empire. The text in the achievement says, "win while not controlling any land cities," which means even when the end-of-game war happens, I will need to consider whether to attack AI land cities.

 

[Tiberiu]

Hi, I think this would be so much easier with Chengsu, because you'd rely on spies to tech and you can stay on fewer cities as a result. I think they are as good as ARC when it comes to this type of strat. The only thing I couldn't say with certainty is what is more profitable in the long run, steal tech or steal science. But I have had good enough results with Steal technology even if the chance is lower but the larger potential reward evens out. But obviously stealing just science is strong...when you steal in 5 turns with so many spies, it becomes piece of cake.

The only real danger is early game, but with a few naval units for defense this is taken care of. Just my 0.02.

 

[vorlon_mi]

Interesting observations about spying. Compared with other games in the franchise, spies in Beyond Earth feel really SLOW. After establishing networks, the first missions to Siphon Energy take 25 turns or more. I assume that the turns decrease as the spies level up, but that levelling up requires multiple missions to be completed. My experience is that it could take 75 turns or more to get them to Special Agent status for a faction that is not Chungsu or ARC. Contrast that with Civ 4, where one only needs to have the agent in the city for 2 or 3 turns to conduct a mission. In Civ 6, fewer turns are needed per mission.
Call me cautious if you will, but I always look at the percentages when I choose a mission. I am very reluctant to choose any missions -- in the mid-game, when I only have 3 spies -- where the agent has < 50% chance of survival. I'm reluctant to risk my investment of 75+ turns on poor odds for Steal Technology. I prefer to keep doing the Siphon Science missions at a higher percentage of success. As a result, I still don't have the Steam achievement for stealing a tech. I've never completed the quest to perform a Covert Op with Intrigue higher than 4.
I agree with @Tiberiu that having multiple agents in the late game, all running Steal Science, is very strong. I have successfully recruited defectors, which was a pleasant surprise. My theory is that for "return on investment", where I'm investing turns to level up my agents, the Steal Science has a better return. Stealing a tech at the cost of a killed agent, and starting all over again, is not a profitable tactic.
But I'm open to persuasion. Are there methods for leveling up faster, or decreasing the chance of agents being killed? What do we know about the game mechanics?

 

[Guide-on]

There is an artifact reward called soul discernment that speeds up spying by 25%. I try to get it if I can. Also on the might tree there is a virtue that speeds up the amount of intrigue. I love ARC.

 

[Tiberiu]

I've never completed the quest to perform a Covert Op with Intrigue higher than 4.

But I'm open to persuasion. Are there methods for leveling up faster, or decreasing the chance of agents being killed? What do we know about the game mechanics?

It's a pity to not get the quest done ASAP, imho. The reward is well worth having a few spies killed. Not only you have one more spy after, but your spies get out already leveled up and are more efficient. (That includes new spies after one dies and respawns)

The way to descrease chance of failure or death is to move the spy away from cities where a counter agent is present. But it's not necessarily the best course of action.

The easiest way to level up faster is to simply get the spying mission done and your spies are always leveled up . Otherwise there is that building quest for surveillance web, if I'm not mistaken, it decreases the times a spy needs to be successful by 1 time per level. Probably worth building one in the empire for the reward, if Culper lodge quest is not completed yet.

Steal Tech is exactly 50% chance for success so it's hit and miss. But given that the reward for costly techs can be 4 times bigger than Steal Science (100% Vs 25%) the ratio of risk/reward remains good enough to be usable. Plus, even if spy dies, usually they are successful and you get a tech in return.
Of course sometimes you can get a cheaper one but it's not a big issue. Another advantage is that the tech is not exactly stolen from anybody (i.e you can get techs that nobody has, this way)

 

[vorlon_mi]

The way to decrease chance of failure or death is to move the spy away from cities where a counter agent is present. But it's not necessarily the best course of action.

Steal Tech is exactly 50% chance for success so it's hit and miss. But given that the reward for costly techs can be 4 times bigger than Steal Science (100% Vs 25%) the ratio of risk/reward remains good enough to be usable. Plus, even if spy dies, usually they are successful and you get a tech in return.
Of course sometimes you can get a cheaper one but it's not a big issue. Another advantage is that the tech is not exactly stolen from anybody (i.e you can get techs that nobody has, this way)

Once my spies have leveled up to Special Agent, yes, I always check to see if a counter agent is present.

You're saying "Steal Tech" will give me a random tech that I don't have -- it could be a cheap one, it could be a big one -- with 50% chance of success. It could be one that I don't care about, it could be the one that I'm 1 turn from finishing, it could be the one that enables my victory wonder. Maybe that works for your style of play, but that seems less valuable to me. The "Siphon Science" mission gives me beakers towards the techs I know that I want.
Again, I go back to the turns required. I don't beeline Computing, since I have several other techs that I want to research first.

• I want to build one Ultrasonic Fence so that the aliens don't destroy my trade routes
• I want to build a couple gunners so that I can kill a siege worm
• Usually I need to unlock a tech or two to develop nearby resources, though these could be done after Computing

Add all that up, with the slower tech rate in the early turns, plus the turns required to build the spy agency, and I don't get any spies until turn 75 or 80. Espionage just isn't -- can't be -- a part of my early game strategy. Mid-game (turns 100-200) yes, I agree.

 

[vorlon_mi]

Postscript: I got the "Terror from the Deep" achievement AND the "Steal a Tech" achievement.

I've begun to reconsider my thoughts above, about "return on investment" by looking at one of the criteria used frequently over in the Civ IV forms: opportunity cost. If I'm not risking my covert agents / spies on high level missions, like "Steal Tech", what am I saving them for? What other uses do they have, especially in the late game as I'm going toward my victory condition?

The only real purpose for a covert agent is to do missions. I can use extra recruits (say, from Feedsite Hubs) to do counter-espionage and keep intrigue down in my cities. The Surveillance Web building helps with that, too. The Black Market (RT) is almost useless, as I have planned my cities to grab resources. As we discussed above, which mission makes the best use of them?

• Siphon Energy is nice, especially when one has activated the Beacon so all excess Energy is being sucked up.
• Recruit Defectors could be good, but my military is pretty strong in the late game. Having a few units to push through the Emancipation Gate (that I didn't have to build myself) is useful, but not amazing.
• Hack Satellites or Sabotage are moderately useful, but only if one wants to conquer the city. Steal Tech is probably better than either of them.

OK, @Tiberiu , you've convinced me. Even though the risk of death is higher, the other missions are also risky and not that much better.